1. Field of the Invention
The present invention relates to a data processing system and a data processing method, particularly for two-dimensional data or image data.
2. Description of the Related Art
When processing 2-dimentional data such as image data in a matrix stored in a virtual storage area, a single row of data is processed piece by piece, in general. After the entire single row of data has been processed, the next single row of data is processed piece by piece in the same manner.
When processing data by hardware that is dedicated to image data processing, the processing is carried out after all the data is stored in an internal memory of the hardware. Accordingly, an image data processing method of dividing image data into a matrix to generate a plurality of blocks is widely used. Each block is executed by data processing, which is carried out block by block independently. In the case where dividing into blocks does not occur, the capacity of the internal memory needs to correspond to the width of an image frame to be executed by data processing. However, in the case where dividing into a plurality of blocks occurs and then data processing in each of the plurality of blocks occurs, the capacity of the internal memory is reduced to a capacity corresponding to the number of data included in each block.
However, when processing data in each block using an algorithm, which is designed without considering such block division processing, the following problems may occur. When processing subject data (hereafter, referred to as ‘subject data’), the results of processing subject data may be considered in another data processing. In this case, an algorithm is designed so that data to be influenced by the results of processing subject data is processed after processing that subject data. However, when processing data in each block, data to be influenced by the results of processing subject data (hereafter, referred to as ‘influenced data’) may be included in a block to be executed by data processing before processing a block including the subject data. In this case, it is difficult to process such influenced data taking the results of processing the subject data into account. With that algorithm, the subsequent block is processed taking into account the influences from a previously processed block. However, it is difficult to reflect in the previously processed blocks the results of processing the subsequent block. Therefore, it is difficult to provide the same results of processing data without dividing into blocks compared to results of processing data with dividing into blocks.
Meanwhile, minute calculation for correction so as to accurately predict an influence on influenced data included in a previously processed block requires an increased amount of calculations. Accordingly, processing speed decreases and/or errors resulting from processing increase.